For the first time, U.S. researchers have integrated ultra-low-noise lasers (ultralow-noise lasers) and photonic waveguides (photonic waveguides) onto a single chip. This long-awaited achievement could make it possible to conduct high-precision experiments using atomic clocks and other quantum technologies in a single integrated device, eliminating the need for room-sized optical platforms in some applications.
Another Leap Forward In Chip Technology
For the first time, U.S. researchers have integrated ultra-low-noise lasers (ultralow-noise lasers) and photonic waveguides (photonic waveguides) onto a single chip. This long-awaited achievement could make it possible to conduct high-precision experiments using
atomic clocks and other quantum technologies in a single integrated device, eliminating the need for room-sized optical platforms in some applications.
When electronics was in its infancy, researchers studied diodes, transistors, and the like as stand-alone devices. The true potential of the technology was not realized until after 1959, when the invention of the integrated circuit made it possible to package all of these
components onto a single chip. Photonics researchers hope to achieve a similar feat of integration, but they face a hurdle: "For photonic links, we need to use a light source (usually a laser) as a transmitter to send the signal to a downstream optical link, such as an
optical fiber or waveguide," explains Chao Xian, a postdoctoral fellow in John Bowers' team at the University of California, Santa Barbara. postdoc Chao Xian explains. "But when you send out light, it usually produces some back reflection: it goes back into the laser and
makes it very unstable."
To avoid such reflections, researchers typically insert isolators. These allow light to pass in only one direction, breaking the natural two-way reciprocity of light propagation. The difficulty is that industry-standard isolators use magnetic fields to accomplish this, which
creates problems for chip fabrication facilities. "CMOS wafer fabs have very stringent requirements for what goes in the cleanroom," explains Xiang, who now works at the University of Hong Kong. "Magnetic materials are usually not allowed."
Integrated But Independent
Because the high temperatures required for annealing waveguides can damage other components, Xiang, Bowers, and colleagues set out to fabricate ultra-low-loss silicon nitride waveguides on silicon substrates. They then covered the waveguides with several layers
of silicon-based material and mounted low-noise indium phosphate lasers on top of the stack. If they had mounted the lasers and waveguides together, the etching involved in fabricating the lasers would have damaged the waveguides, but bonding subsequent layers
on top avoided this problem.
Separating the laser from the waveguide also means that the two devices interact with each other by means of an intermediate silicon nitride "redistribution layer" that passes through their evanescent field (the component of the electromagnetic field that does not
propagate but decays exponentially away from the source). Thus, the distance between them minimizes unwanted interference. "The top laser and the bottom ultra-low-loss waveguide are far apart," Xiang says, "so they can both perform independently. The control of
the silicon nitride redistribution layer allows them to be coupled exactly where you want them. Without it, they wouldn't be coupled together."
Combined active and passive devices
The researchers showed that this laser device is robust to noise to the level expected in standard experiments. They also demonstrated the practicality of their device by tuning the beat frequency between two such lasers to produce a tunable microwave frequency
generator - something that was previously impractical on integrated circuits.
Given the widespread use of ultra-low-noise lasers in modern technology, the team says that being able to use such lasers in integrated silicon photonics is a huge leap forward. ",On the same chip, we can combine the active and the passive components," Xiang said.
"The next step is to use these ultra-low-noise lasers for very complex optical functions, such as precision metrology and sensing."
Scott Diddams, an optical physicist at the University of Colorado at Boulder who was not involved in the research, was impressed: "The problem of integrated lasers with optical isolators has been a focus of industry attention for at least a decade, and no one has
figured out how to solve the problem of making really low-noise on-chip lasers on a chip ...... so this is a real breakthrough," he says. "People like John Powers have been working in this field for 20 years, so they know the basic building blocks, but figuring out how to
get them to work together perfectly isn't just like bolting the pieces together."
Diddams adds that new integrated devices could be "very influential" in the quantum computing space. "Serious companies are trying to build platforms that involve atoms and ions - which run in very specific colors - and we talk to them with lasers," he explained.
"Without integrated photonics like this, one cannot build a fully functional quantum computer on a large scale."